Efficiently Reconstructing Dynamic Scenes One D4RT at a Time

Chuhan Zhang, Guillaume Le Moing, Skanda Koppula, Ignacio Rocco, Liliane Momeni, Junyu Xie, Shuyang Sun, Gabriel Brostow, Rahul Sukthankar, Joëlle K. Barral, Raia Hadsell, Zoubin Ghahramani, Andrew Zisserman, Junlin Zhang, Mehdi S. M. Sajjadi

Abstract:

Understanding and reconstructing the complex geometry and motion of dynamic scenes from video remains a
formidable challenge in computer vision. This paper introduces D4RT, a simple yet powerful feedforward model
designed to efficiently solve this task. D4RT utilizes a unified transformer architecture to jointly infer depth, spatiotemporal correspondence, and full camera parameters from a single video. Its core innovation is a novel querying mechanism that sidesteps the heavy computation of dense, perframe decoding and the complexity of managing multiple, task-specific decoders. Our decoding interface allows the model to independently and flexibly probe the 3D position of any point in space and time. The result is a lightweight and highly scalable method that enables remarkably efficient training and inference. We demonstrate that our approach sets a new state of the art, outperforming previous methods across a wide spectrum of 4D reconstruction tasks.